Research On Identification And Optimization Of Abnormal Batteries In Lithium Battery Pack

Lithium battery has the advantages of high power density,no memory effect,and long service life,so it has broad application prospects in military,transportation,power stations and other fields.In order to meet different power requirements,single cells are usually connected in series and parallel to form a battery pack.Locating and processing the abnormal attenuated batteries in the battery pack is the key to ensure the high performance and safe output of the energy storage system,and to avoid failure and even danger.Therefore,this thesis studies the abnormal attenuation battery in the battery pack from three aspects:identification,state estimation and maintenance decision,aiming to systematically solve the problem of accurate location,estimation and maintenance of abnormal battery in the battery pack.First of all,this thesis builds the attenuation model of a single cell from the perspectives of heat,electricity,and attenuation coupling.The electrical characteristic model adopts the2-RC equivalent circuit model.The thermal characteristic model considers the environmental temperature and the heating loss of the battery itself.The attenuation characteristics consider the impact of charge and discharge current,depth of discharge and temperature.Then we build a 3*3 small battery pack model based on the attenuation model of the single cell.Then,under different working conditions,study the influence of different degrees of attenuation of batteries in different positions in the battery pack on the overall thermal and electrical characteristics of the battery pack,and compare with normal battery packs,the characteristic variable combinations representing the abnormal attenuation battery are analyzed and obtained.After that,the characteristic variable data sets containing normal and abnormal batteries are combined with the least square support vector machine(LS＿SVM)algorithm for training.A classifier which takes the characteristic variable as input and whether the battery has abnormal attenuation as output is obtained.In order to obtain and process the specific state data of abnormal batteries,a battery state assessment model is built in this thesis.By obtaining the characteristic variable data set of batteries at different attenuation rates,and combined with BP neural network,the regression model with characteristic variables as input and battery attenuation rate as output was obtained.Then the simulation test is carried out,the test results show that the established classifier and regression model are accurate and effective,and the accuracy rate is more than 95%.Finally,in order to choose a suitable maintenance strategy,three schemes of removal,replacement and balance are analyzed from the perspectives of economy and treatment effect.The results show that the balanced maintenance method has the best maintenance effect but the highest economic cost,which leads to the worst economy.The maintenance method of replacing the new battery has good effect,and the battery scale is small,which leads to the cost of the new battery is too large,and the economy is poor.At the same time,there is a shutdown loss when replacing the battery.The maintenance method of excised battery prolongs the service life appropriately and has high economy,but there are unstable factors.By comparing the life and economy of different strategies,it provides theoretical support for the high efficiency application of lithium battery pack.In this thesis,based on the extraction of characteristic variables representing abnormal batteries in the battery pack,the identification,location,state estimation and maintenance decision of abnormal batteries are systematically studied by combining with relevant algorithms.It provides ideas and reference methods for realizing intelligent diagnosis,intelligent prediction and accurate positioning of lithium energy storage system.